From owner-freebsd-hackers Mon Feb 28 7: 5:40 2000 Delivered-To: freebsd-hackers@freebsd.org Received: from search.sparks.net (search.sparks.net [208.5.188.60]) by hub.freebsd.org (Postfix) with ESMTP id B1ECE37B75F for ; Mon, 28 Feb 2000 07:05:36 -0800 (PST) (envelope-from dmiller@search.sparks.net) Received: by search.sparks.net (Postfix, from userid 100) id 78893DBB4; Mon, 28 Feb 2000 10:05:30 -0500 (EST) Received: from localhost (localhost [127.0.0.1]) by search.sparks.net (Postfix) with ESMTP id 644DADBB3; Mon, 28 Feb 2000 10:05:30 -0500 (EST) Date: Mon, 28 Feb 2000 10:05:30 -0500 (EST) From: David Miller To: Joerg Micheel Cc: freebsd-hackers@freebsd.org Subject: Re: Need help with crash debug In-Reply-To: <20000228171931.B24598@cs.waikato.ac.nz> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG On Mon, 28 Feb 2000, Joerg Micheel wrote: > David, > > I haven't done too much debugging with FreeBSD either. A few small > notes, however. > > I don't think the gdb version shipped is ready to handle SMP kernels. > This is a deficiency, not sure someone is working on it, I believe not. > Wow. So if smp core dumps can't be debugged post mortem, how do the developers develop smp kernels? > Having said this, it is probably impossible to find the kernel stack > pointer (as there are several of them anyways). Which in turn will make > it impossible to detect where exactly the kernel crashed. Unless you > have seen the machine crashing plus the report on the console, there > is not much you can hope for with this crash dump. > > If this error is repeatable, you may have luck. Disable SMP in the > kernel and reboot, let the machine running until it crashes again. > Then, have another try. I may resort to this. If it's an smp related bug, this will just make it "go away", which would be OK, except there are times when I need the extra horsepower. > FWIW, I do have a book on kernel crash dump analysis, it is for Solaris, > but I expect a great deal to be identical to general kernel debugging, > I haven't touched it in years. My debugging is somewhat different these > days. > > Hope this helps. Thanks for the input:) --- David To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message